1/X circuit

[Tony Allgood]

>how accurate do you need? what you want is called a divider & you can
make an approximate one
with anything that you make a multiplier with (eg a 3080).

This sounds good... and I will have a go today. It doesn't need to be
too accurate... not very good engineering term that...

To be honest, I don't know how to even specify how good I need it. What
I am trying to do is derive an error term for a voltage controlled wave
symmetry VCO, by directly scaling the CV to the VCO. ie. modulate saws
into triangles into inverted saws and back again. I can do it,
reasonably well, but the frequency is shifting too much as I increase
modulation depth. This is no suprise, I expected that, but I need to
scale the modulating signal by 1/x to get zero frequency shift. At
present I have a linear VCO and a CV scaling VCA driven from the
modulating signal. This will always produce a drop in frequency as the
mark space ratio increases because rate at which the shorter ramps are
decreasing is not proportional to the rate at which the longer ramps are
incresing. Korg's MS20s LFO managed it but it was controlled by pots not
voltages.

Back to the drawing board...

Regards,

Tony Allgood  Penrith, Cumbria, UK